Using Numpy To Convert An Int To An Array Of Bits

I need a way to convert 20 million 32 and 64-bit integers into corresponding bit arrays (so this has to be memory/time efficient). Under advice from a different question/answer her

Solution 1:

Try: dtype='>i8', like so:

In [6]: np.unpackbits(np.array([1], dtype='>i8').view(np.uint8))
Out[6]: 
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1], dtype=uint8)

Reference:

http://docs.scipy.org/doc/numpy/user/basics.byteswapping.html